DESCRIPTION Abstract: Eighty to ninety percent of what most young children learn about the world comes through vision. The same cannot be said when we seek to learn about the inner workings of our own body, because light beyond ?skin deep? becomes diffused due to multiple scattering. Instead, researchers have resorted to alternative means?such as X-ray, magnetic resonance, and ultrasound?to probe deep into the body. Until now, most advances in optical imaging have been geared towards high-resolution functional and molecular imaging at depths less than 1 mm in scattering tissue. The pursuit of deep-tissue optical imaging with high spatial resolution has been stymied by the inherent optical diffusion?the grand challenge since the inception of biomedical optics. We must meet this challenge to reach the full potential of light because it is such a powerful tool from both the physical and biological perspectives. Physically, the tiny fraction of the electromagnetic spectrum that light covers is the only part that probes molecular structures directly; biologically, the ability of molecules to sense, react to, and emit light is encoded on the most fundamental (i.e., genetic) level! In addition, light as nonionizing radiation is as safe to biological organisms as air and water. Therefore, light is the most natural choice for visualizing biological structures and events, interrogating and controlling biological processes, as well as diagnosing and treating diseases, if only we could overcome the optical diffusion?a seemingly unbreakable barrier. While multiple scattering of light is treated as a problem in conventional wisdom, I believe that it should be part of the solution. Our recent work on time-reversed ultrasonically encoded (TRUE) optical focusing (Nature Photonics 2011) is a first breakthrough in this direction. TRUE focusing can noninvasively deliver light to a dynamically defined focus deep in a scattering medium. This invention opens the door to an even greater paradigm-shifting opportunity?one that controls the photon paths to minimize transmission loss in tissue. Here I propose a novel technology, called ?photon tunneling?, to achieve such an audacious goal. Photon tunneling aims to send light deep into biological tissue along dynamically ?drilled? light tunnels. Unprecedented light penetration depth, limited by only absorption instead of orders-of-magnitude stronger scattering, can be reached. Because the optical absorption coefficient is as low as 0.1/cm, the 1/e penetration can be as deep as 10 cm (~4 inches). If successfully developed, such a groundbreaking technology would revolutionize biomedicine. Applications can be found in all aspects of biomedical optics, including imaging (e.g., fluorescence tomography and reporter gene imaging), sensing (oximetry and glucometry), manipulation (optogenetics and nerve stimulation), and therapy (photodynamic therapy and photothermal therapy). An NIH Director?s Pioneer Award would grant me the intellectual freedom and resources to develop a completely new field that will literally illuminate the core of biomedical research.